Disk 123456789123456789 is UNAVAIL

[Spud]

I'm sure I've missed something here but anyway, I get this email telling me

New alerts:
* Pool Tank1 state is DEGRADED: One or more devices could not be opened.
Sufficient replicas exist for the pool to continue functioning in a degraded
state.
The following devices are not healthy:

* Disk 123456789123456789 is UNAVAIL

How do I work out which HD is the problem from that number (which I have changed obviously)

And wouldn't it be easier to just say ada0?

Thanks in advance.

 

[NugentS]

The issue is that ada0/1/2/3 etc can change when you reboot depending on luck and which drive comes up first - and yes its a nusiance

So to find out which disk - and I notice you haven't posted your hardware as per forum rules.......

IF the disk is still recognised by TrueNAS go to storage disks (or wherever it is in the TN version you have) and find the serial number of the disk then search physically in the machine for that disk with that serial number

If however the disk is so broken that its not recognised then you will need to note all the serial numbers of the recognised disks and find the one thats missing in a physical search

 

[Spud]

The issue is that ada0/1/2/3 etc can change when you reboot depending on luck and which drive comes up first - and yes its a nusiance

So to find out which disk - and I notice you haven't posted your hardware as per forum rules.......

IF the disk is still recognised by TrueNAS go to storage disks (or wherever it is in the TN version you have) and find the serial number of the disk then search physically in the machine for that disk with that serial number

If however the disk is so broken that its not recognised then you will need to note all the serial numbers of the recognised disks and find the one thats missing in a physical search

My hardware consists of a Supermicro X10SL7-F motherboard with an IBM M1015 controller and 32G of memory and the CPU is a Xeon E3-1241.

I didn't think any of that was relevant to be honest, and still don't.

I'm running the latest version of TrueNAS-13.0-U6.1

The drive is still in the system and now working fine, the problem pops its head up after a reboot sometimes and I think its the sata cable which I want to replace. But with 18 drives I'm having a little trouble finding out which ones the problem.

I don't have any problems finding serial numbers, in fact I use an app on my phone to read the barcodes thats easy.

My problem is finding the drive from the email which refers to the drive as "123456789123456789" and trying to work out what drive its talking about.

Thanks for the help!

 

[Spud]

The issue is that ada0/1/2/3 etc can change when you reboot depending on luck and which drive comes up first - and yes its a nusiance

I always thought that say da0 and so no was assigned to the controller it was connected to on the MB

So learnt some new today :)

 

[Patrick M. Hausen]

My problem is finding the drive from the email which refers to the drive as "123456789123456789" and trying to work out what drive its talking about.

Go to the UI > Storage > Disks

 

[sretalla]

My problem is finding the drive from the email which refers to the drive as "123456789123456789" and trying to work out what drive its talking about.

To be more deterministic, if the disks are still online, you can get that from:

zdb -l /dev/da0p2 (p2 since the Pool will typically consume the second partition with swap as the first)...

That should spit out a bunch of stuff about the pool the disk is a member of, including the ZFS identifier or guid of that disk.

You would still possibly need to then find the serial number of that disk to find it physically in the chassis, so smartctl -i /dev/da0 | grep Serial would get you that

 

[Spud]

Go to the UI > Storage > Disks

Yeah only thing in there is the serial number. Not what I'm looking for.

 

[Redcoat]

Use glabel status

in combo with Patrick's Storage>Disks

 

[Spud]

To be more deterministic, if the disks are still online, you can get that from:

zdb -l /dev/da0p2 (p2 since the Pool will typically consume the second partition with swap as the first)...

That should spit out a bunch of stuff about the pool the disk is a member of, including the ZFS identifier or guid of that disk.

You would still possibly need to then find the serial number of that disk to find it physically in the chassis, so smartctl -i /dev/da0 | grep Serial would get you that

zdb -l /dev/da0p2 | grep guid | less

That only gives me 14 different guids for that one hard drive alone. So 18 HDs x 14 guids per disc is 252 lines to search through

But I see where your coming from, thanks it just might take longer then expected.

Thanks again!

 

[Spud]

Use glabel status

View attachment 74325

in combo with Patrick's Storage>Disks

Yeah thats not the guid I'm look for dude!

 

[Redcoat]

Yeah thats not the guid I'm look for dude!

Oh, after some research - I now "see" what you are after.

 

[sretalla]

I think what you're needing in that other case is the output from zpool status, which would show the gptids to match glabel status

 

[Spud]

I think what you're needing in that other case is the output from zpool status, which would show the gptids to match glabel status

Yeah I can't see how that would help me to be honest but all good.

I'm going to turn off these notifications as they do a great job in telling you theres a problem, but a not so good job in telling you exactly where the problem is.

Thanks again!